Animal Molecular Breeding 2024, Vol.14, No.5, 307-317 http://animalscipublisher.com/index.php/amb 313 statistical power. Combining datasets from different populations or conducting meta-analyses can help increase sample sizes and improve the robustness of GWAS findings (Gebreyesus et al., 2019; Bakhshalizadeh et al., 2021). Despite these strategies, the risk of false positives persists, necessitating rigorous validation of identified SNPs in independent populations (Pryce et al., 2010). 7.3 Challenges in identifying causative variants Identifying causative variants among the numerous associated SNPs remains a daunting task. Many GWAS identify regions of the genome associated with traits, but pinpointing the exact causative mutations requires further fine-mapping and functional studies. For example, while some studies have successfully narrowed down genomic intervals containing causative mutations using haplotypes, the precision of quantitative trait loci (QTL) mapping can still be limited (Pryce et al., 2010). Longitudinal GWAS and the use of whole-genome sequence data can provide more detailed insights, but they also require high-quality imputation and extensive computational resources (Teng et al., 2023). Moreover, the polygenic nature of milk production traits, where multiple genes contribute small effects, complicates the identification of individual causative variants (Buaban et al., 2021). 8 Integrating GWAS with Other Genomic Tools 8.1 Use of genomic selection and marker-assisted selection (MAS) Genomic selection (GS) and marker-assisted selection (MAS) have revolutionized dairy cattle breeding by leveraging the power of genome-wide association studies (GWAS). GS uses genomic breeding values (GEBV) calculated from dense genetic markers across the genome, capturing the effects of quantitative trait loci (QTL) that contribute to variation in traits such as milk production. This method has shown significant improvements in the reliability of breeding values, leading to faster genetic gains compared to traditional selection methods (Hayes et al., 2009). MAS, on the other hand, focuses on specific markers associated with desirable traits identified through GWAS. For instance, the identification of SNP markers and haplotypes associated with milk production traits has been validated across different breeds, enhancing the precision of QTL mapping and the effectiveness of MAS (Pryce et al., 2010; Gutierrez-Reinoso et al., 2021). The integration of these genomic tools into breeding programs has not only increased productivity but also helped mitigate issues like inbreeding depression by providing more accurate selection criteria (Gutierrez-Reinoso et al., 2021). 8.2 Combining GWAS with epigenomics and metabolomics The integration of GWAS with epigenomics and metabolomics offers a comprehensive approach to understanding the genetic architecture of milk production traits. Epigenomics studies the modifications on DNA and histones that affect gene expression without altering the DNA sequence, providing insights into gene regulation mechanisms. Metabolomics, which involves the large-scale study of small molecules (metabolites) within cells, tissues, or organisms, can reveal the biochemical pathways influenced by genetic variations. Combining these approaches with GWAS can identify not only the genetic variants associated with milk production but also how these variants influence metabolic pathways and gene expression. For example, studies have identified candidate genes and biological networks related to milk production and somatic cell score (SCS), implicating pathways such as intracellular cell transportation and protein catabolism (Buaban et al., 2021). This holistic approach can lead to the discovery of novel genes and pathways, offering new targets for genetic improvement and better understanding of the underlying biology of milk production traits (Marina et al., 2021). 8.3 Future prospects of integrating GWAS with machine learning techniques The future of integrating GWAS with machine learning (ML) techniques holds great promise for enhancing the accuracy and efficiency of genomic predictions. Machine learning algorithms can handle large and complex datasets, making them ideal for analyzing the vast amount of data generated by GWAS. These techniques can identify complex patterns and interactions between genetic variants that traditional statistical methods might miss. For instance, ML can improve the prediction accuracy of GEBV by incorporating non-linear relationships and interactions among SNPs (Hayes et al., 2009). Additionally, ML can be used to refine the selection of candidate variants for genomic evaluation, as demonstrated by studies that have used meta-analysis and joint analyses to select the most relevant SNPs for milk production traits (Berg et al., 2016; Teissier et al., 2018). The integration of
RkJQdWJsaXNoZXIy MjQ4ODYzNA==